Fine-Tuning AI

Chapter 4. Excavating Data

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 4th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

Model architectures are usually very open. There are reference implementations that are open source, papers that are available to read and replicate, or tutorials showing you how to code from scratch. Similarly, training techniques, hyperparameters, and best practices are easy to find! But data, particularly the data that you or your customers have gleaned from years of practicing in your business domain is proprietary, domain-specific, and hard won. It’s the thing that differentiates you from everyone around you.

It’s the data moat that surrounds your opportunity and distinguishes you from others.

One thing to note, and we’ll focus this on during this chapter, is the data within your data. Often, we have datasets that have already been structured for us in order to be able to build and use computer-based systems. But, in the age of AI, and with the artificial understanding that modern models can give you, you also have the opportunity to find value within the massive ...